Afonso S. Bandeira's Homepage



						401-4944-20L Mathematics of Data Science (Spring 2020, ETHZ) Afonso S. Bandeira Lecture will be online, I will post a zoom link on Piazza 10min before the start of the class. Lectures: Tue&Thu 15-17 (Tue: HG F 7, Thu: HG G 3). Office Hours: Afonso's office hours posted weekly on calendar in here. Federica and Christoph office hours: Tuesday's 2-3pm, alternating between the two: March 17th: HG G 22.1 (Federica), March 24th: HG G 21.3 (Christoph), and continuing with this pattern. Teaching Assistants: Federica Cecchetto and Christoph Glanzer Piazza page available here. Announcements: Extra Problem Set available here There is now a full set of lecture notes here: CLN. See also this Book Draft: BD - the lecture notes are essentially Chapters 2-10 of the Book Draft. Please sign up to the piazza page here for announcements. When feasible use Piazza to ask questions, myself and the TAs will make an effort to answer questions quickly on Piazza. This course will be similar to a courses I have given before, you can see more info e.g. at MIT OCW here or these Lecture Notes: Ten Lectures and Forty-Two Open Problems in the Mathematics of Data Science. You can find more info at ETH VVZ. Open problems will be presented at the end of most lectures. I am here to help: If you have any question, want to discuss a problem, or brainstorm about any research idea, just stop by office hours, write me on piazza, or email me and we'll schedule a time to meet. Feedback: Also, if you have any comment or feedback on the class (it's going too fast, too slow, you want me to cover more of something, or less of something else, etc) please let me know, submit a comment to this google form (or in person, or through email). Having direct feedback from you is the best way for me to try give lectures that you like! (keep in mind that I don't know who sent me the comment or feedback and there is no way for me to answer, for questions use piazza or email, piazza tends to receive replies faster). Weekly Syllabus (we skipped what I originally called ``Chapter 7'', the Chapters are now numbered sequentially 2 through 10 in CLN.): Feb 18: Introduction to the course and Syllabus Feb 20: Probability and Surprises in High Dimensions. Chapter 2 of CLN or BD. Also Section 4.1. of 10L42P. Feb 25: Continuation. Chapter 2 of CLN or BD. Basics of Linear Algebra. Chapter 3 of CLN or BD. Also Section 0.3. of 10L42P. Feb 27: Principal Component Analysis. Chapter 3 of CLN or BD. Also Sections 1.1. and 1.2 of 10L42P. Mar 03: Marcenko-Pastus and BBP Transition. Chapter 3 of CLN. Also Section 1.3 of 10L42P. Mar 05: Continuation of Marcenko-Pastus and BBP Transition. Some open problems in Section 1 of 10L42P. Mar 10: Page Rank. Chapter 4 of CLN or BD, see also here. Mar 12: Introduction to Spectral Graph Theory, Chapter 4 of CLN or BD. See recording below. Also Section 2 of 10L42P Mar 17: Normalized Cut, Spectral Clustering, Cheeger's inequality. End of Chapter 4 of CLN or BD. Also Section 3 of 10L42P Mar 19: Diffusion Maps. Nonlinear dimension reduction. Chapter 5 of CLN or BD. Also Section 2 of 10L42P Mar 24: Semi-Supervised Learning. Chapter 5 of CLN or BD. See recording below. Also Section 2 of 10L42P Mar 31: Gaussian Concentration and Matrix Concentration Chapter 6 of CLN or BD (also Sections 4.2 and 4.4 of 10L42P) Apr 2: Continuation Matrix Concentration Chapter 6 of CLN or BD (Sections 4.5 and 4.6 of 10L42P) Apr 7: End of Matrix Concentration Chapter 6 of CLN or BD (rest of section 4.6 of 10L42P). (very) Brief start of Max-Cut Chapter 7 of CLN or BD (Section 7 of 10L42P) Apr 9: Max-Cut and Approximation Algorithms Chapter 7 of CLN or BD (Section 8 of 10L42P) Apr 21, 23, and 28: Community Detection in the SBM Chapter 8 of CLN or BD (Section 9 of 10L42P) Apr 30: Exercise session for Problem Set 2 (see Problem Set below) May 5, 7,12: Dimension reduction via random projections Chapter 9 of CLN or BD (Section 5 of 10L42P) May 17: Compressed Sensing Chapter 10 of CLN or BD (Section 6 of 10L42P) May 19: Exercise session for Problem Set 3 (see Problem Set below) May 26: Compressed Sensing Chapter 10 of CLN or BD (Section 6 of 10L42P) May 28: Compressed Sensing Chapter 10 of CLN or BD (Section 6 of 10L42P) and final remarks Lecture Recordings (``redos'', some done from home due to the COVID-19 Pandemic): Mar 10: part 1 and part 2 Mar 12: full Mar 17: part 1 and part 2 Mar 19: part 1 and part 2 Mar 24: part 1 and part 2 Mar 31: part 1 and part 2 Apr 2: part 1 and part 2 Apr 7: part 1 and part 2 Apr 9: part 1 and part 2 Apr 21: full Apr 23: part 1 and part 2 Apr 28: part 1 and part 2 May 5: full May 7: full May 12: part 1 and part 2 May 14: full May 26: part 1 and part 2 May 28: full and extra Lecture Notes: Course Lecture Notes: CLN (contains some material that was not covered in lecture, but mentioned and is natural further reading; note that Chapter numbering is now sequential, 2 through 10.) Book Draft: BD The Lecture notes correspond, essentially, to Chapters 2-10 of the book draft. My co-authors and I will keep updating the book draft, adding material, improving figures, fixing typos, etc. *Many thanks for all your comments on the notes/book!* 10L42P: Ten Lectures and Forty-Two Open Problems in the Mathematics of Data Science Problem Sets: (solutions will be posted on Piazza) Problem set 1 Problem set 2 Problem set 3 Extra Problem Set 1 Open Problems: List available at the course lecture notes.